A5006156483- Topic Modeling
- Natural Language Processing Techniques
- Artificial Intelligence in Healthcare and Education
- Machine Learning in Healthcare
- Biomedical Text Mining and Ontologies
- Schizophrenia research and treatment
- Multimodal Machine Learning Applications
- Mental Health Treatment and Access
- Ethics in Clinical Research
- Cancer Genomics and Diagnostics
- Electronic Health Records Systems
- Biomedical and Engineering Education
- Healthcare cost, quality, practices
- Semantic Web and Ontologies
- Advanced Text Analysis Techniques
- Colorectal Cancer Screening and Detection
- Health Sciences Research and Education
- Cultural Heritage Management and Preservation
- Sentiment Analysis and Opinion Mining
- Artificial Intelligence in Healthcare
- AI in cancer detection
- Ethics and Social Impacts of AI
- Historical Education Studies Worldwide
- Quality Function Deployment in Product Design
- Radiomics and Machine Learning in Medical Imaging
Duke Institute for Health Innovation
2023-2024
Massachusetts Institute of Technology
2020-2022
Radboud University Nijmegen
2020
Stanford University
2020
University of Arizona
2020
University of Massachusetts Lowell
2015-2016
Contextual word embedding models such as ELMo and BERT have dramatically improved performance for many natural language processing (NLP) tasks in recent months. However, these been minimally explored on specialty corpora, clinical text; moreover, the domain, no publicly-available pre-trained yet exist. In this work, we address need by exploring releasing text: one generic text another discharge summaries specifically. We demonstrate that using a domain-specific model yields improvements 3/5...
Private and public sector structures norms refine how emerging technology is used in practice. In healthcare, despite a proliferation of AI adoption, the organizational governance (i.e. institutional governance) surrounding its use integration often poorly understood. What Health Partnership (HAIP) aims to do this research better define requirements for adequate systems healthcare settings support health system leaders make more informed decisions around adoption. To work towards...
Abstract When integrating AI tools in healthcare settings, complex interactions between technologies and primary users are not always fully understood or visible. This deficient ambiguous understanding hampers attempts by organizations to adopt AI/ML, it also creates new challenges for researchers identify opportunities simplifying adoption developing best practices the use of AI-based solutions. Our study fills this gap documenting process designing, building, maintaining an solution called...
In recent years, there has been a growing interest in the field of "AI Ethics" and related areas. This is purposefully broad, allowing for intersection numerous subfields disciplines. However, lot work this area thus far centered computational methods, leading to narrow lens where technical tools are framed as solutions broader sociotechnical problems. work, we discuss less-explored mode what it can mean "do" AI Ethics: tech worker collective action. Through action, employees powerful...
Machine learning has been suggested as a means of identifying individuals at greatest risk for hospital readmission, including psychiatric readmission. We sought to compare the performance predictive models that use interpretable representations derived via topic modeling human experts and nonexperts. examined all 5076 admissions general psychiatry inpatient unit between 2009 2016 using electronic health records. developed multiple predict 180-day readmission these based on features from...
This paper describes TwitterHawk, a system for sentiment analysis of tweets which participated in the SemEval-2015 Task 10, Subtasks A through D. The performed competitively, most notably placing 1 st topicbased classification (Subtask C) and ranking 4 th out 40 identifying sarcastic tweets.Our submissions all four subtasks used supervised learning approach to perform three-way assign positive, negative, or neutral labels.Our development efforts focused on text pre-processing feature...
This paper describes the SimiHawk system submission from UMass Lowell for core Semantic Textual Similarity task at SemEval-2016.We built four systems: a small featurebased that leverages word alignment and machine translation quality evaluation metrics, two end-to-end LSTM-based systems, an ensemble system.The LSTMbased systems used either simple LSTM architecture or Tree-LSTM structure.We found of three base feature-based model obtained best results, outperforming each model's correlation...
Precise evaluation metrics are important for assessing progress in high-level language generation tasks such as machine translation or image captioning.Historically, these have been evaluated using correlation with human judgment.However, human-derived scores often alarmingly inconsistent and also limited their ability to identify precise areas of weakness.In this paper, we perform a case study metric by measuring the effect that systematic sentence transformations (e.g.active passive voice)...
The recent release of many Chest X-Ray datasets has prompted a lot interest in radiology report generation. To date, this been framed as an image captioning task, where the machine takes RGB input and generates 2-3 sentence summary findings output. quality these reports canonically measured using metrics from NLP community for language generation such Machine Translation Summarization. However, evaluation (e.g. BLEU, CIDEr) are inappropriate medical domain, clinical correctness is critical....
Private and public sector structures norms refine how emerging technology is used in practice. In healthcare, despite a proliferation of AI adoption, the organizational governance surrounding its use integration often poorly understood. What Health Partnership (HAIP) aims to do this research better define requirements for adequate systems healthcare settings support health system leaders make more informed decisions around adoption. To work towards understanding, we first identify standards...